Three child characteristics thought to decelerate or prevent spoken language development in some children with ASD that are especially difficult to measure are (a) slow speech processing, (b) atypical word processing, and (c) atypical pre-speech vocalizations. The proposed project seeks to improve the measurement of these characteristics and to address the following research questions: (a) Do speed of speech processing, typicality of word processing, and complexity of vocalizations (i.e., the three new putative predictors of language) have positive, predictive relations with later productive or receptive lexical levels in initially nonverbal preschoolers with ASD? (b) Are the concurrent correlations among the three new putative predictors of language positive and significant in children with ASD? (c) Can the relation between speed of speech processing or typicality of word processing and later lexical levels be explained by covariation with attention to child-directed speech in initially nonverbal children with ASD? (d) Can the relation between complexity of vocalizations and later productive language be explained by covariation with gross measures of vocal communication or oral-motor ability in initially nonverbal children with ASD? The supplement will use an existing R01's longitudinal correlational design and a subgroup of the children with ASD that are already being recruited in the parent R01 grant to address the aims. The supplemental funds will be used to (a) recruit a new sample of 35 typically developing (TD) infants, (b) assess event-related potentials [ERPs] in the TD sample and the subgroup of the R01's larger ASD sample, and (c) assess complexity of vocalizations in the ASD subgroup. ERPs to speech sounds or words (and their contrast stimuli) will be assessed in 40 preschoolers (24 - 48 months) with ASD who are initially nonverbal at R01-study entry and 35 typically developing (TD) infants who are matched on lexical comprehension (approximately 8 - 18 mos.) to measure speed of speech processing or typicality of word processing. Based on our past rates of attaining "good" ERP data in these populations, expected analyzable sample size for the ERP data is 30 in each sample. A novel use of TD infants'ERP data will be implemented to derive measures of individual differences in processing in the ASD group. In the subgroup of ASD participants, day-long audio-recorded samples of children's vocalizations will be collected to derive estimates of the developmental level of ASD children's vocalizations using a very recently-developed acoustic analysis method (i.e., 1st peer-reviewed use was published in PNAS on 7/10/2010, Oller et al., 2010). The work will improve our understanding of the variance in early spoken language development in initially nonverbal preschoolers with ASD and the pattern of intercorrelation among selected putative predictors of language that may be unique to the ASD population. This understanding will eventually provide empirical guidance in selecting among existing communication treatments and help us develop new communication treatments for nonverbal children with autism. PUBLIC HEALTH RELEVANCE: This revision is responding to PA-10-158 (Research on Autism and Autism Spectrum Disorders), which is intended to support research goals for the Strategic Plan for ASD. The strategic plan for ASD research contains a long-term goal of understanding the heterogeneity of ASD. The proposed research will improve our understanding of the variance in early spoken language development in initially nonverbal preschoolers with ASD, set-up future research to provide empirical guidance in selecting among existing communication treatments, and generate data that will stimulate future development of new communication treatments for nonverbal children with autism.